Contents

What's Inside

Chapter One · Facts checked June 2026

What Kids Are Actually Doing with AI

Let's start with the deep breath: your child is almost certainly already using AI, they almost certainly started before you knew, and — here's the part the headlines skip — most of what they're doing with it is perfectly sensible. This chapter replaces assumptions with data, because good parenting decisions need accurate pictures.

The actual picture

The Australian research (the eSafety Commissioner's 2026 report among it) paints a consistent scene for children aged 10–17: most are already using AI, with homework help leading the list, followed by creative projects, general curiosity questions — and, for a meaningful minority, digital companionship (more on that in Chapter Eight, calmly). Usage starts younger than most parents expect and grows steadily through the teens. Explore the data yourself in the AU Youth AI Data Explorer; the assumption quiz that goes with it is humbling for nearly everyone, me included.

The perception gap

Here's the finding that should change your week: research from Common Sense Media, Pew and others shows parents and children see AI use very differently — parents consistently underestimate how often their kids use AI, how early they started, and how emotionally significant some of that use is. The gap matters because you can't guide what you can't see, and children don't volunteer what they suspect will be banned. The fix isn't surveillance; it's making AI a normal dinner-table topic rather than a confession.

Myths worth retiring

A quick spring-clean of the beliefs that derail family conversations: AI isn't conscious and doesn't "know" your child (it predicts plausible text — that's the whole trick); using AI isn't automatically cheating (it depends entirely on how — Chapter Seven has the full traffic-light system); AI doesn't always tell the truth (it errs confidently, which is why verification is a family habit, not a nice-to-have); and banning AI doesn't protect children (it just relocates their use somewhere you can't see it). The myth-vs-fact game covers the lot — play it with your kids.

Chapter Two · Facts checked June 2026

The Three Buckets & Choosing Safe Platforms

Not all AI is created equal, and the single most useful mental model in this entire book is a simple sorting rule. Master it and you'll know, within seconds of hearing about any new app, what kind of conversation your family needs to have.

The three buckets

Every AI tool your child encounters fits one of three buckets:

• Learning-first AI — tutoring apps, school platforms, educational tools. Purpose-built for kids, usually well-guarded. Needs light-touch guidance: an interested parent, occasional check-ins.
• General AI — ChatGPT, Claude, Gemini and friends. Powerful, built for adults, enormously useful for homework and creativity. Needs active guidance: age-appropriate settings, ground rules, and the verification habits of Section II.
• Companion AI — apps designed to be a friend, character or confidant. This bucket carries the highest emotional stakes and needs genuine attention: clear boundaries, open conversation, and your presence. (Chapter Eight is devoted to it.)

The bucket determines the conversation. A new app appears in your child's life? First question, always: which bucket? The sorting game trains the reflex in ten minutes.

Comparing platforms like a pro

Within any bucket, platforms differ enormously on the things parents care about: age restrictions (and whether they're enforced or merely stated), parental controls, content filtering, what happens to your child's data, and cost. The safety comparator puts six popular platforms side by side, and the recommender matches platforms to your child's age and your family's priorities. Two rules of thumb travel well: prefer platforms with real parental controls over platforms with mere promises, and check the data practices before the features — a dazzling app that trains its models on your child's conversations is not a bargain.

Chapter Three

Ages 4–11: First Concepts & Guided Discovery

AI guidance, like everything in parenting, is a moving target that grows up alongside your child. This chapter covers the early years — where the goal isn't AI use at all, but AI understanding, built in ways small children actually enjoy.

Ages 4–7: concepts without screens

Here's the lovely surprise: the best AI education for little ones involves no screens whatsoever. AI's core ideas — sorting by rules, learning from examples, making predictions, getting things wrong — are all teachable at the kitchen table. The course's six offline activity cards turn them into games: sort the toy animals by a secret rule and let your child guess it (that's classification); play "robot commands" where instructions must be exact (that's programming logic); guess what comes next in a pattern (that's prediction). The explanation script that goes with the sorting game is perfect five-year-old pitch: "You just did what a computer does — you learned the rule from examples!" The aim at this age: AI is a machine that spots patterns, made by people, and sometimes wrong. That's the whole curriculum, and it's a brilliant one.

Ages 8–11: guided discovery

Now screens enter, with you alongside. Two skills define this stage. First, choosing the right tool for the job — the course's search-vs-generate exercise teaches the distinction that even adults fumble: facts need verified sources (search), creative and learning tasks suit AI (generate), and personal or emotional questions belong with a real person, full stop. Children who learn to pick the right source early build critical thinking that lasts a lifetime.

Second, writing prompts that actually work — the prompt ladder turns vague asks into good ones step by step: "tell me about dogs" becomes "explain why dogs wag their tails, for a nine-year-old, in three sentences". It's fun, it works immediately, and it quietly teaches the deepest lesson in this whole book: the quality of the answer depends on the quality of the question.

Chapter Four

Ages 12–18: Verification, Disclosure & Independence

The teenage years flip the script: your role shifts from gatekeeper to coach, and the skills shift from supervision to self-management. Done well, your eighteen-year-old leaves home with something genuinely valuable — professional-grade AI judgement.

Ages 12–15: the sceptical years

Perfect timing, really — adolescence is when scepticism comes free. Two trainable skills anchor this stage. First, auditing AI homework help: AI answers look convincing while containing subtle errors, vague hedging and outright fabrications, and the Homework Helper Audit trains teens to mark each claim as correct, vague or wrong. The habit it builds — never submit what you haven't checked — is worth more than any single grade. Second, spotting synthetic media: deepfakes, manipulated images and AI-generated content are part of their information diet now; the Deepfake Detective exercises train the tells. A teen who can catch AI's mistakes has learned something better than rules: judgement.

Ages 16–18: workflows and honest independence

Senior students are ready to use AI as the genuine productivity tool it is — and to learn the professional skill that keeps it honest: disclosure. The workflow designer helps them decide which steps of an assignment AI may assist (research summaries, outline feedback, proofreading) and which stay manual (the thinking, the writing, the conclusions). The disclosure template builder then formats an honest statement of what was used and how. Frame it the way the course does: disclosure isn't confession — it's the professional standard they'll use at university and in every job thereafter. Seniors who disclose well are ahead of most adults.

The family decision flowchart

Tying every age together: the Family AI Decision Flowchart — a simple decision tree to run before any AI interaction (What's the task? Which bucket? What would the school say? Who checks the output?), with a printable, customisable version for the fridge. The course's wisdom on this deserves italics: the best AI rules are ones your family creates together. Children who help set the boundaries actually follow them — a principle that has survived every technology since the television.

Chapter Five · Facts checked June 2026

Safety & Privacy: Controls, Data & the Law

Now the practical armour. None of this chapter requires technical skill — just an evening, a cup of tea, and the willingness to open some settings menus you've been avoiding.

Parental controls, platform by platform

Every major AI platform now offers some form of parental controls — age-appropriate modes, content filters, usage visibility — and almost nobody switches them on. The setup wizard walks each platform step by step with tick-off checklists and a family safety score. The priority system is sensible: critical items (age settings, content filters) first, important ones (data settings, history controls) next, nice-to-haves after. One honest caveat the course makes plain: no parental control replaces conversation — settings are the seatbelt, not the driving lesson.

What the tools collect

Here's the part that surprises most parents: every conversation your child has with an AI tool is potentially stored — and may be used to train future models. Platforms differ widely on what they collect, how long they keep it, and whether children's data trains their systems; the data explorer breaks each one down in plain English, with quick-action guides for opting out of training and deleting history. Two family rules cover most of the risk: real names, addresses, school names and photos stay out of AI chats; and opt out of data training wherever the setting exists. Five minutes of settings, lasting protection.

The Australian legal picture

For Australian families, the law is moving in your favour: the Privacy Act review and the incoming Children's Online Privacy Code are raising the bar for how platforms may collect and use children's data, alongside the eSafety Commissioner's existing powers. You don't need to become a lawyer — the 5-Point Family Action Card turns the legal landscape into five concrete steps, printable for the fridge. Knowing your child's rights makes you a more confident advocate when a platform behaves badly — and platforms are noticeably politer to parents who cite the Code.

Chapter Six

Red Flags & the Family Safety Audit

This is the chapter I hope you never need — and the one worth reading most carefully, calmly, in advance. Knowing the warning signs before you need them is the difference between a wobble caught early and a problem discovered late.

The eight warning signs

The course identifies eight patterns suggesting a child's relationship with AI may be heading somewhere unhealthy, each rated amber (monitor closely) or red (act now). The ambers: rising secrecy about AI use, preferring AI conversation to family time, sleep disruption from late-night sessions, declining schoolwork alongside heavy AI use. The reds: emotional dependence on an AI companion ("it's the only one who understands me"), withdrawing from real friendships, distress when access is removed, and seeking comfort from AI for serious emotional struggles. The recognition cards pair every sign with what to look for, how to respond, and — crucially — a conversation opener that doesn't begin with an accusation.

Responding without the explosion

The pattern across every recommended response: curiosity before consequences. "I've noticed you're chatting with that app a lot — what do you like about it?" opens a door; "give me your phone" closes every door in the house. Most ambers resolve with attention and adjusted boundaries. The reds deserve seriousness: genuine emotional dependence on AI, or AI standing in for help with real distress, is a moment for professional support — your GP, a school counsellor, or Kids Helpline (1800 55 1800). Asking for help early isn't overreaction; it's good parenting.

The family safety audit

Then make checking a routine rather than a panic response. The family safety audit walks through every platform, setting, device and household rule in about twenty minutes, and produces a printable report card with prioritised actions. The honest checklist of sixteen questions about the past month is private — never stored or sent — and worth answering truthfully. Platforms update, kids download, settings drift: re-audit each school term and the drift never gets far.

Chapter Seven

Homework & Honesty: The Traffic Light System

The question every parent eventually asks: "Is my child cheating?" The honest answer: it depends entirely on how AI is being used — and the good news is that the line is teachable, and most kids genuinely want to know where it is.

Green, amber, red

The course's traffic-light system sorts every homework scenario into three zones. Green — AI is fine: explaining a concept differently, generating practice questions, checking grammar, brainstorming ideas. Amber — AI with disclosure: research summaries, outline feedback, improving a draft the child wrote — fine, provided it's declared. Red — not appropriate: AI writing the essay, solving the assessment, doing the thinking the task exists to develop. The scenario sorter trains the instinct across twenty real cases, and the printable traffic-light poster belongs on the study wall. The tie-breaker question that resolves edge cases: what is this homework trying to teach — and is AI doing that part?

Verification: the family habit

Chapter Four's audit skill deserves its place in daily homework life, because trust without verification is the #1 mistake students make with AI. The three-step trust checker keeps it simple enough for school nights: Can I find this claim in a real source? Does it match what I learned in class? Would I bet a mark on it? The practice lab — six AI answers seeded with plausible errors — makes a surprisingly competitive family game. A child who checks AI's work is learning the most transferable skill in this book.

Creativity: brainstorm, don't copy

For creative work, the line is ownership: AI may spark ideas, suggest directions, critique drafts — but the child stays in the driver's seat and does the actual making. The boundary explorer maps the spectrum from "your work" to "AI's work" scenario by scenario, and the story-with-guardrails exercise proves the happy medium: AI offers three story starters, your child writes every word that follows. Pride of authorship survives; blank-page terror doesn't.

Schools and disclosure

Finally, align home rules with school rules. The policy decoder translates your school's AI policy into the three zones above (and if your school hasn't got one, the ten ready-made questions will politely start that conversation). Then make disclosure normal: the statement generator formats honest declarations from footnote to full process note. The framing that lands with kids: disclosure isn't admitting weakness — it's what professionals do, and examiners reward honesty far more often than they punish it.

Chapter Eight

Wellbeing, Screen Time & Companion AI

The chapter for the worries that keep parents up at night — handled, I promise, without alarmism. The research offers genuinely reassuring frameworks; the companion AI conversation needs care rather than panic; and the whole thing ends with a routine your family can actually keep.

Beyond counting minutes: the 5 Cs

The paediatric 5 Cs framework (adapted from the American Academy of Pediatrics) replaces minute-counting with questions that actually predict wellbeing: Child (how does this particular child respond?), Content (what are they actually doing?), Calm (is tech used to regulate emotions?), Crowding out (what is it replacing — sleep, sport, friends?), and Communication (do they talk about their online life?). The liberating insight from the quality rater: it's not about how long — it's about what they're doing. An hour of creative AI projects and an hour of passive scrolling are not the same hour.

Companion AI: the honest conversation

Children are forming relationships with AI companions — designed to be warm, endlessly available and never disappointed — and dismissing this with "it's just a chatbot" misses why it appeals: the companion never judges, never tires, never has a bad day. Real friends do, and that's precisely what makes them real. The risk explorer works through six realistic scenarios; the boundary cards script the conversations, including what not to say ("that's pathetic" closes the door forever) and what to say instead:

And the line that must stay bright: feelings of sadness, anxiety or loneliness deserve a real person — a parent, a counsellor, a helpline — never just an algorithm trained to keep the conversation going. The when-to-talk-to-a-real-person flowchart makes that judgement teachable.

Building the healthy routine

Wellbeing is ultimately built from boring, repeated structure: purpose rules (AI for creating and learning beats AI for filling silence), time boundaries (no AI in the hour before bed — sleep is non-negotiable), and device-free zones (dinner table, bedrooms — for everyone, parents included; children spot hypocrisy at forty paces). The routine planner assembles your family's version and feeds it straight into the agreement waiting in Chapter Nine.

Chapter Nine

Your Family AI Agreement & the Monthly Review

Everything in this book now condenses into two family rituals: one document, signed together, and one conversation, repeated monthly. Between them, they turn eight chapters of guidance into the way your household actually runs.

The family AI agreement

The agreement builder assembles your family's rules across six steps: family details, approved tools per child (Chapter Two's buckets, applied), usage rules per context (homework via the traffic light, creative work via brainstorm-don't-copy, entertainment within the routine), privacy rules (Chapter Five's non-negotiables), time and place boundaries (Chapter Eight's routine), and what happens when rules are missed — which should emphasise conversation and recalibration over punishment, because the goal is judgement, not compliance. Then the crucial bit of theatre: print it and sign it together, parents included, with parents' own commitments on the page. A signed agreement on the fridge has authority no lecture ever achieves.

The monthly review

Rules without review become wallpaper within a term. The monthly review template structures a twenty-minute family check-in across five sections: what AI did we use this month? Did anything concern anyone (every family member gets a voice)? Are the safety settings still right? What's new in the AI world? And what shall we adjust? The companion journal tracks the family's confidence growing month by month — genuinely encouraging to look back on. Keep the tone of a family meeting about pizza toppings, not a disciplinary hearing: light, regular, everyone heard.

Chapter Ten · Facts checked June 2026

Resources, the Literacy Assessment & Where Next

The final chapter sends you off well-stocked: free resources for every age, a way to measure how far your family has come, and an honest word about the road ahead.

The free shelf

You need not spend a cent to raise an AI-literate family. The resource directory curates the good free stuff — filterable by age and type — and the two 8-week family curricula (younger and older children) turn it into a structured plan: one topic a week, activities included, printable in full, startable this weekend. The suggested rhythm is the right one: explore first, pick one thing, do it together, then build from momentum rather than ambition.

Measuring how far you've come

The family literacy assessment rates your household across four dimensions — understanding, safety, school use and family communication — and the progress tracker tallies the course's skills towards your AI-Literate Family certificate. A certificate on the fridge is pleasingly silly and genuinely motivating in roughly equal measure; children, in my experience, take certification far more seriously than legislation.

An honest send-off

The tools your children use next year don't exist yet, and that's fine — because nothing in this book depended on any particular app. The buckets, the traffic light, the verification habit, the boundaries, the agreement, the monthly conversation: every one of them transfers to whatever arrives next. You haven't learned to police a technology; you've built a family culture of curiosity, honesty and healthy scepticism — which, you may notice, is just good parenting wearing a new hat. Your kids are growing up in the most remarkable technological moment in generations, and they're lucky: they've got a parent who bothered to read the manual. Now go and ask them what they made today.